As the issue of climate change worsens due to carbon emissions, extensive research has been conducted on sustainable energy sources to replace fossil fuels. Among them, hydrogen production through water electrolysis has garnered attention as a future energy source because it allows electricity generated from renewable sources to be stored in the form of chemical energy without emitting carbon dioxide. However, the oxygen evolution reaction (OER) occurring at the anode during electrolysis has been identified as a bottleneck due to its inherently high activation energy required for four-electron transfer, prompting discussions on catalyst development based on electronic and crystal structures to improve hydrogen production efficiency. While reducing overpotential to achieve high catalytic activity has been a primary concern in electrolysis over the past decade, recent emphasis has shifted towards the durability and structural stability of catalysts for prolonged reactions. As either a strong alkaline electrolyte or an acidic membrane is adopted under a highly oxidizing environment, understandings of corrosion resistance, lattice oxygen participation, elemental leaching, and subsequent amorphization at the catalyst surfaces in harsh pH conditions have been major research directions. In other words, to design high-durability catalysts, systematic investigation of surface changes in OER catalysts as a function of overpotential is essential. For atomic-scale tracking of the catalyst surface varying with overpotential, both the selection of a model catalyst material and the specific form of the material are important. First, for crystallographically identical surface analysis, a (001) heteroepitaxial thin film of perovskite oxide catalyst was adopted. The thin film model system enables systematic observation of surface changes during electrochemical reactions, allowing easy tracking of structural changes at the atomic level during the OER reaction in perovskites utilizing catalytically inactive A-site and active B-site transition metals, which can be readily traced via STEM. Second, based on the pourbaix diagram, intentional selection of A-site and B-site elements was performed. Alkali and alkaline-earth metals (Li, Na, K, Ca, Sr, and Ba), which represent A-site cations, have the characteristic of easily dissolving in ionic form in a wide pH range under anodic potential. Therefore, chemically stable La, without dissolution, was chosen as the A-site element. For the B-site, Co was selected as an OER active element, excluding elements that are either catalytically inactive (Ti, Mn, and Fe) or prone to leaching (V, Cr, and Ni) in the electrolyte. Thus, in this study, LaCoO3 was adopted as the thin film model system. Using heteroepitaxial (001) LaCoO3 thin films (thickness: 20nm) in this work, we demonstrate the structural and compositional changes of the film surface as a function of the anodic overpotential at an atomic level. In particular, through a combination of direct STEM imaging, image simulations, and surface analyses together with density functional theory (DFT) calculations, we clarify that the surface degradation behavior varies, strongly depending on the applied overpotential during the OER. When conducting the electrochemical reaction through a 2-hour chronoamperometry (CA) at 1.65V (vs. a reversible hydrogen electrode (RHE), IR-corrected), the current tends to degrade continuously. Through BF-STEM imaging, a significant weakening of contrast at the topmost surface was observed at 1.65V, with simulations revealing that atomic displacements on the angstrom scale are the cause. The atomic displacements at the surface subsequently lead to a reduced density of the Co 3d states near the Fermi level (EF), which are the major origins of catalytic degradation without notable surface restructuring. In contrast, at potentials higher than 1.67V, a rapid degradation occurs within 30 minutes to 1 hour, followed by a stabilization or even an increase in current. At the same time, substantial surface amorphization occurred at 1.67V. Despite claims that lattice oxygen in LaCoO3 does not contribute to OER catalysis, SIMS analysis confirmed significant lattice oxygen exchange (isotope 18 oxygen-water signal) at higher overpotential conditions (≥1.67V), leading to the formation of amorphous hydroxide on the film surface. The presence of an OH peak in XPS and changes in the O K-edge in EELS analysis supported this. In this process, despite the thermodynamic stability of Co, the oxygen octahedral structure surrounding Co collapses due to lattice oxygen evolution, leading to Co dissolution into the electrolyte and exhibiting Co-deficient features on the surface. Ultimately, this study demonstrated through various STEM images and surface analyses that the catalytic reaction mechanism differs significantly depending on the overpotential. By successfully visualizing the surface changes of oxide catalysts at the atomic level, our research highlights the importance of overpotential as a critical factor in determining the stability of catalysts during OER. Figure 1
Read full abstract